Gangliogliomas of the Spinal Cord: Neuroimaging Correlations with Pathology, Controversies in Pathological Diagnosis, and Prognosis

Abstract

Gangliogliomas of the spinal cord are not, in general, different from those in the brain, but there are special problems in their neuroradiological and pathological diagnosis. While often stated to be quite rare, the available literature suggests that these are from 10% to 25% of all intramedullary spinal cord neoplasms in children, and are less frequently found in adults. Their neuropathological diagnosis is highly dependent on adequate biopsies or resections, most often accomplished at centers with pediatric neurosurgeons accustomed to aggressive approaches to intramedullary spinal cord tumors. Neuroradiologically these are intrinsic cord masses which are on average longer, in terms of spinal segments, than other intramedullary tumor types; many are “holocord” tumors. The tumors usually enhance even though most are low grade. The neuropathological diagnosis as in the brain depends on finding a tumor with cells that include both glial cells, generally astrocytic, and large neurons or ganglion cells. Proof that neuron-like cells in a tumor are tumor ganglion cells involves abnormal H&E appearances and immunohistochemical stains, of which the most important is that for synaptophysin. The problems and controversies of interpretation of synaptophysin immunostains in the spinal cord, in view of somewhat similar patterns of immunopositivity on anterior horn cells and tumor ganglion cells, is illustrated and discussed. Other helpful techniques include stains for myelin and for axons. These are low grade tumors with a favorable oncologic prognosis if successfully gross totally excised.